A Query Relevant Context Driven Ontology Recommendation System incorporating Semantics Preservation and Semantic Ontology Matching

Abstract

The World Wide Web is evolving into a standard Semantic Web that requires efficient modeling of
knowledge bases. Knowledge Bases are the constructed mainly based on the domain level segregation of Ontologies.
Most interestingly, dynamic construction of knowledge bases is a vital and an important task wherein query relevant
domain level ontology bases are constructed. Ontology Recommendation is a methodology to construct knowledge bases
and is vital in the context of Semantic Web. It is quite important to retain the initial associations and axioms between the
ontologies as they are recommended to preserve ontology semantics. A semantic strategy that preserves associations
among the ontology entities during recommendation of ontologies has been proposed. In this approach, domain level
OWL ontologies are converted into RDF by the derivation of intermediate XML parse trees. A HashMap-HashTable
methodology is used to preserve the axioms between the ontological concepts and individuals. The SemantoSim measure
for computing the semantic similarity has been proposed. The semantic relatedness is computed between the query and
the concepts at first and then between the query and the description logics which makes this a context driven ontology
recommendation system. The context based ontology recommendation system with ontology relationship preservation
yields an overall accuracy of 86.87 %.